Consider the data given in Table 10-7 (on the textbook's Web site) relating to stock prices and GDP for the period 1980-2006.
a. Estimate the OLS regression
Yt = B1 + B2Xt + ut
b. Find out if there is first-order autocorrelation in the data on the basis of the d statistic.
c. If there is, use the d value to estimate the autocorrelation parameter ρ.
d. Using this estimate of p, transform the data per the generalized difference equation (10.14), and estimate this equation by OLS (1) by dropping the first observation and (2) by including the first observation.
e. Repeat part (d), but estimate ρ from the residuals as shown in Eq. (10.20). Using this estimate of p, estimate the generalized difference equation (10.14).
f. Use the first difference method to transform the model into Eq. (10.17) and estimate the transformed model.
g. Compare the results of regressions obtained in parts (d), (e), and (f). What conclusions can you draw? Is there autocorrelation in the transformed regressions? How do you know?